This thesis is an investigation into the accuracy of hybrid density functional theory to predict the properties of two transition metal oxides: Ilmenite (FeTiO3) and haematite (sigma-Fe2O3). The hybrid density functional theory examined is Becke's B3LYP functional, which is an empirical mix of density functional theory and exact nonlocal exchange from Hartree-Fock theory. For bulk ilmenite, results from the B3LYP functional are compared with Hartree-Fock and pure density functional theory calculations. The computed properties are found to be very sensitive to the treatment of electronic exchange and correlation, with the best results being achieved using the hybrid functional. Calculations performed using the hybrid functional benefit from its better treatment of the electronic self interaction and its reasonable estimate of the pair correlation energy of the doubly occupied Fe-d orbital. To assess the performance of the hybrid functional in simulating Fe2O3 and FeTiO3 with different cation-anion coordination than that found in ilmenite or haematite, studies were performed on their high pressure polymorphs, for which there are a range of experimental results for comparison. This tests the transferability of the functional before examining cases, such as the surfaces of these materials, where there are little or no experimental or theoretical results. For the currently known high pressure polymorphs of ilmenite and haematite, the structural and elastic parameters computed using the hybrid functional are found to be in good agreement with those observed, as is the predicted stability of the phases. In ilmenite, the calculations predict the stability of a new high-pressure polymorph with space group Cmcm, occurring at pressures above 44 GPa. Calculations of the high pressure polymorphs of haematite involve the examination of a range of charge, spin, and magnetic states for each of the polymorphs. Magnetic ordering was found to be important for all the polymorphs, and for each polymorph an antiferromagnetic ordering was found to be lower in energy than the ferromagnetic ordering. The predicted transition pressure from the corundum structure and the magnetic collapse of the Fe3+ cations were in good agreement with experiment. At high pressures the lowest energy configuration for the orthorhombic perovskite structure was computed to occur with mixed high-spin /low-spin Fe3+ cations, in contrast to predictions in the literature of a Fe2+/Fe4+ solution. The CaIrO3-type structure was also computed to be stable with a mixed high-spin/ low-spin Fe3+ configuration at high pressures, and is computed to be the most stable polymorph at pressures above 46 GPa at 0 K. The structure of the ilmenite (0001) surface is examined using the B3LYP functional, and for this surface twelve different terminations are considered, with surface energies and relaxed geometries calculated. The Fe terminated (0001) surface was found to have the lowest cleavage energy, and also to be the most stable surface at low oxygen partial pressures suggesting it is most likely to form when ilmenite is cleaved under high vacuum.

No description available.

No description available.

In this thesis we use density functional theory (DFT) to study the solvent mediated interactions between solvophobic, solvophilic and patchy nanostructures namely rectangular cross section blocks. We calculate both the density profiles and local compressibility around the blocks and the results obtained for our model system provide a means to understanding the basic physics of solvent mediated interactions between nanostructures, and between objects such as proteins in water, that possess hydrophobic and hydrophilic patches. Our results give an improved understanding of the behaviour of liquids around solvophobic objects and solvophobicity (hydrophobicity) in general. Secondly, we look into the physics incorporated in standard mean-field DFT. This is normally derived by making what appears to be a rather drastic approximation for the two body density distribution function: ρ(2)(r,r′) ≈ ρ(r)ρ(r′), where ρ(r) is the one-body density distribution function. We provide a rationale for why the DFT often does better than this approximation would make you expect. Finally, we develop a lattice model to understand the nature of the pattern formation exhibited by certain systems of particles deposited on liquid-air interfaces and in particular the nature of the transitions between the different patterned structures that are observed. This is done using Monte Carlo computer simulations and DFT and links the observed microphase ordering with the micellisation process seen e.g. in surfactant systems.

Le développement d'un nouveau médicament est un processus long et coûteux. Entre la détermination d'une cible thérapeutique et la mise sur le marché d'un nouveau médicament, plus de dix ans de recherche sont nécessaires pour un coût supérieur à un milliard d'euros.L'accélération de ce processus et la réduction de son coût restent un enjeu majeur. Pour y parvenir, les simulations numériques, peu coûteuses et rapides, sont massivement utilisées. Malgré cela, elles restent limitées, en partie à cause de la quantité très importante de molécules de solvant à considérer. La théorie de la fonctionnelle de la densité moléculaire permet d'étudier la solvatation de composés de n'importe quelle taille et de n'importe quelle forme. Elle prédit en quelques secondes seulement à la fois l'énergie libre de solvatation et une carte détaillée de la densité d'équilibre autour de ce soluté. Ces grandeurs étant à la base de nombreux autres calculs utilisés par l'industrie pharmaceutique, la MDFT ouvre donc une autre voie d'optimisation de ces process. Cette thèse consiste à effectuer le premier pas vers l'ensemble de ces applications. Pour cela, nous avons adapté la théorie ainsi que le code associé avant de l'appliquer à des systèmes biologiques. / Drug development is time and cost-consuming: It takes in average 10 years and 1 billion euros to move from a therapeutic target to a new drug. To speed up this process and reduce its cost, numerical simulation are massively used. Nevertheless, they remain limited, one reason of which is the huge amount of solvent molecules to consider. The molecular density functional theory is a liquid state theory that allows the study of the solvation thermodynamics of solutes of arbitrary shape. MDFT predicts, in few seconds only, the free energy of solvation and the solvent profils. These parameters are at the heart of many others calculation used by the pharmaceutical industry. This thesis is the first step towards these applications. For that purpose, we adapted the theory as well as the associated code to this new target, then applied them to system of biological interest.

abstract: This thesis focuses on the theoretical work done to determine thermodynamic properties of a chalcopyrite thin-film material for use as a photovoltaic material in a tandem device. The material of main focus here is ZnGeAs2, which was chosen for the relative abundance of constituents, favorable photovoltaic properties, and good lattice matching with ZnSnP2, the other component in this tandem device. This work is divided into two main chapters, which will cover: calculations and method to determine the formation energy and abundance of native point defects, and a model to calculate the vapor pressure over a ternary material from first-principles. The purpose of this work is to guide experimental work being done in tandem to synthesize ZnGeAs2 in thin-film form with high enough quality such that it can be used as a photovoltaic. Since properties of photovoltaic depend greatly on defect concentrations and film quality, a theoretical understanding of how laboratory conditions affect these properties is very valuable. The work done here is from first-principles and utilizes density functional theory using the local density approximation. Results from the native point defect study show that the zinc vacancy (VZn) and the germanium antisite (GeZn) are the more prominent defects; which most likely produce non-stoichiometric films. The vapor pressure model for a ternary system is validated using known vapor pressure for monatomic and binary test systems. With a valid ternary system vapor pressure model, results show there is a kinetic barrier to decomposition for ZnGeAs2. / Dissertation/Thesis / M.S. Materials Science and Engineering 2011

Dans le domaine des nanosciences, la recherche sur les matériaux possédant des dimensions réduites a connu des progrès spectaculaires. Tandis que de nombreux travaux ont été fait initialement sur le graphène, l'attention s'est ensuite portée vers d'autres matériaux bidimensionnels, tels que le nitrure de bore hexagonal ou encore les dichalcogénures de métaux de transition. Néanmoins, il est toujours nécessaire de trouver des matériaux possédant des caractéristiques équivalentes ou supérieures à celles des composés déjà connus. Dans le cadre de cette thèse, nous avons utilisé le calcul ab initio et plus particulièrement la théorie de la fonctionnelle de la densité pour prédire et comprendre les propriétés de trois familles de matériaux bidimensionnels. Premièrement, en prenant la structure du phosphorène comme structure de référence et en remplaçant le phosphore par des atomes voisins dans le tableau périodique, nous avons pu obtenir des matériaux inconnus jusqu'ici. Ensuite, nous nous sommes intéressés à des matériaux à base d'halogénures tels que AcOBr ou BaFCl, parmi d'autres. Enfin, nous avons mis l'accent sur des composés bidimensionnels quaternaires, tels que ScP2AgSe6, P2AgSe6Bi, P2CuBiSe6 et CuInP2 S6. Pour chaque matériau, nous avons démontré qu'il était dynamiquement stable et étudié sa structure électronique, et pour certains l'effet d'un champ électrique sur le matériau, ce qui ouvre la porte à de futures études expérimentales dans le domaine / In the field of nanosciences, research on materials with reduced dimensions has seen spectacular progress. While many works were initially done on graphene, the attention then came to other two-dimensional materials, such as hexagonal boron nitride or transition metal dichalcogenides. Nevertheless, it is still necessary to find materials with characteristics equivalent to or superior to those of the already known compounds. In this thesis, we used ab initio calculations and more particularly density functional theory to predict and understand the properties of three families of two-dimensional materials. First, taking the phosphorene structure as the reference and replacing phosphorus with neighboring atoms in the periodic table, we have been able to obtain unknown materials so far. Then we looked at halide materials such as AcOBr or BaFCl, among others. Finally, we have focused on two-dimensional quaternary compounds, such as ScP2AgSe6, P2AgSe6Bi, P2CuBiSe6 and CuInP2S6. For each compound, we demonstrated that it was dynamically stable and studied its electronic structure, and for some the effect of an electric field on the material, which opens the door for future experimental studies in the field

Mixed ionic and electronic conductor oxides, in particular La1-xSrxCoyFe1-yO3-d (LSCF), have been widely used as the cathode materials in solid oxide fuel cells for high-temperature energy applications. The focus of this thesis is primarily on constructing the instability phase diagram of Sr segregations on LSCF surfaces at the experimentally relevant temperatures and oxygen partial pressures using the first-principles density functional theory (DFT). A generic first-principles free-energy functional is developed to obtain the nonstoichiometric oxygen vacancy concentrations for the bulk and surface phases. These results agree well with the corresponding thermo-gravimetry measurements, and furthermore suggest that the oxygen vacancies are energetically stabilized at surfaces for all temperatures and oxygen partial pressures, while such surface stabilization effects become stronger at higher temperatures and lower oxygen partial pressures. Based on these nonstoichiometric oxygen vacancy predictions, we construct the free-energy phase diagrams of the Sr-segregation reaction as a function of temperature, oxygen partial pressure, and CO2 partial pressure for both the bulk and surface LSCF phases. Our results suggest that Sr segregations strongly accumulate towards the LSCF surface phase where the oxygen vacancy nonstoichiometries are abundant. Our results also indicate that the Sr segregation reactions are significantly enhanced at high temperatures, low oxygen partial pressures, and high CO2 partial pressures. The computed reaction temperature ranges are consistent with the total reflection X-ray fluorescence (TXRF) measurements.

Research on two-dimensional (2D) materials currently occupies a sizeable fraction of the materials science community, which has led to the development of a comprehensive body of knowledge on such layered structures. However, the goal of this thesis is to deepen the understanding of the comparatively unknown heterostructures composed of different stacked layers. First, we utilise linear-scaling density functional theory (LS-DFT) to simulate intricate interfaces between the most promising layered materials, such as transition metal dichalcogenides (TMDC) or black phosphorus (BP) and hexagonal boron nitride (hBN). We show that hBN can protect BP from external influences, while also preventing the band-gap reduction in BP stacks, and enabling the use of BP heterostructures as tunnelling field effect transistors. Moreover, our simulations of the electronic structure of TMDC interfaces have reproduced photoemission spectroscopy observations, and have also provided an explanation for the coexistence of commensurate and incommensurate phases within the same crystal. Secondly, we have developed new functionality to be used in the future study of 2D heterostructures, in the form of a linear-response phonon formalism for LS-DFT. As part of its implementation, we have solved multiple implementation and theoretical issues through the use of novel algorithms.

This thesis presents work done on the calculation of the effects of anharmonic nuclear motion on the properties of solid materials from first principles. Such anharmonic effects can be significant in many cases. A vibrational self-consistent field (VSCF) method is used as the basis for these calculations, which is then improved and applied to a variety of solid state systems. Firstly, work done to improve the efficiency of the VSCF method is presented. The standard VSCF method involves using density functional theory (DFT) to map the Born-Oppenheimer (BO) energy surface that the nuclei move in, a computationally expensive process. It is shown that the accurate forces available in plane-wave basis DFT can be used to help map the BO surface more accurately and reduce the computational cost. This improved VSCF+f method is tested on molecular and solid hydrogen, as well as lithium and zirconium, and is found to give a speed-up of up to 40%. The VSCF method is then applied to two different systems of physical interest. It is first applied to the case of the neutral vacancy in diamond, in order to resolve a known discrepancy between harmonic ab initio calculations and experiment -- the former predict a static Jahn-Teller distortion, whilst the latter leads to a dynamic Jahn-Teller effect. By including anharmonic corrections to the energy and nuclear wavefunction, we show that the inclusion of these effects results in agreement between first-principles calculations and experiment for the first time. Lastly, the VSCF method is applied to barium titanate, a prototypical ferroelectric material which undergoes a series of phase transitions from around 400 K downwards. The nature of these phase transitions is still unclear, and understanding them is an active area of research. We describe the physics of the phase transitions of barium titanate, including both anharmonicity and the effect of polarisation caused by long wavelength vibrations, to help understand the important physics from first principles.