Computational Studies of Inorganic Systems with a Multiscale Modeling Approach: From Atomistic to Continuum Scale

Multiscale modeling is an effective tool for integrating different computational methods, creating a way of modeling diverse chemical and physical phenomena. Presented are studies on a variety of chemical problems at different computational scales and also the combination of different computational methods to study a single phenomenon. The methods used encompass density functional theory (DFT), molecular dynamics (MD) simulations and finite element analysis (FEA). The DFT studies were conducted both on the molecular level and using plane-wave methods. The particular topics studied using DFT are the rational catalyst design of complexes for C—H bond activation, oxidation of nickel surfaces and the calculation of interaction properties of carbon dioxide containing systems directed towards carbon dioxide sequestration studies. Second and third row (typically precious metals) transition metal complexes are known to possess certain electronic features that define their structure and reactivity, and which are usually not observed in their first-row (base metal) congeners. Can these electronic features be conferred onto first-row transition metals with the aid of non-innocent and/or very high-field ligands? Using DFT, the impact of these electronic features upon methane C—H bond activation was modeled using the dipyridylazaallyl (smif) supporting ligand for late, first-row transition metal (M) imide, oxo and carbene complexes (M = Fe, Co, Ni, Cu; E = O, NMe, CMe2). To promote a greater understanding of the process and nature of metal passivation, first-principles analysis of partially oxidized Ni(111) and Ni(311) surface and ultra-thin film NiO layers on Ni(111) was performed. A bimodal theoretical strategy that considers the oxidation process using either a fixed GGA functional for the description of all atoms in the system, or a perturbation approach, that perturbs the electronic structure of various Ni atoms in contact with oxygen by application of the GGA+U technique was applied. Binding energy of oxygen to the nickel surfaces, charge states of nickel and oxygen, and the preferred binding mode of oxygen to nickel were studied to gain a better understanding of the formation of oxide layers. Using density functional theory, the thermodynamic properties for developing interaction potentials for molecular dynamics simulations of carbon dioxide systems were calculated. The interactions considered are Ni + H2O, Ni + Ni, Ni + CO2, CO2 + CO2, CO2 + H2O and H2O + H2O. These systems were chosen as the possible interactions that can occur when carbon dioxide is stored in the ocean. Molecular dynamics simulations using the results from the DFT studies were also conducted. Finally, thermal conduction analysis was performed on layered functionally graded materials (FGM) subjected to thermal shock by sudden cooling of the material in order to investigate the results obtained from three different mixing laws: linear, quadratic, and half-order. The functionally graded material considered was a composite of nickel and carbon nanotubes at different compositions varying from two to five layers. The middle layers for the three to five layers are composed of graded (i.e., gradually changing) percentages of nickel and carbon nanotube. The thermal conductivity, specific heat and density for the composites were calculated depending on the percentages of materials in each layer, and assuming different rules of mixture