With the passage of time and the advancement of our industrial civilization, environmental concerns have become more and more recognized since the 1990s. Carbon dioxide reduction reactions are capable of converting carbon dioxide into valuable hydrocarbons and reducing the carbon emission from the combustion of fossil fuels. This is a promising direction for sustainable energy resources given that the scarcity of fossil fuels is becoming more threatening to the survival of mankind. In recent years, oxide-derived metal nanostructures have been synthesized and show unique catalytic features. Recently, Sloan et al. synthesized a novel oxide-derived copper nanocube structure, which showed a high selectivity toward ethylene over methane and low overpotentials. In this work, the presence of subsurface oxygen in the catalyst surface is tested with density functional theory (DFT) calculations, as a complement to experimental x-ray photoelectron spectroscopy. Due to limitations on the scale of modeling with DFT, the results indicate a very low stability of subsurface oxygen, which give rise to a question if subsurface oxygen would be stable with a reasonably large cluster model. Self-consistent charge density functional tight binding (SCC-DFTB) is adopted to investigate a nanocube model. In this model, a manually reduced cuprious oxide nanocube is constructed and investigated. Subsurface oxygen atoms close to facets are found to be more stable inside. A higher degree of disorder is proposed to be the cause of this difference in stabilizing subsurface oxygen atoms between the slab and nanocube models. The presence of subsurface oxygen enhances the adsorption of CO on the Cu(100) surface, increasing the likelihood for adsorbed CO molecules to dimerize, which is the rate determining step for ethylene production on Cu(100) under low-overpotential conditions. With subsurface electronegative atoms such as oxygen or fluorine, it is also found that the d-band scaling relation could be broken.

The carbon dioxide reduction reaction is a promising candidate to tackle ecological challenges of our age. This is due to its capability of reducing carbon dioxide emission generated from the combustion of fossil fuels by converting carbon dioxide into valuable hydrocarbons. Oxide-derived metal nanostructures have been found to exhibit unique catalytic characteristics for facilitating the carbon dioxide reduction reaction. In this thesis work, the stability, influence, and effects of subsurface oxygen atoms are investigated by theoretical computations with various levels of theory and models. It is found that subsurface oxygen atoms are stable and that their presence increases the CO adsorption strength and coverage on oxide-derived Cu surface. This is explained by a reduced σ-repulsion and leads to the breaking of scaling relations. Although it does not directly reduce the CO dimerization barrier, the adsorption of H atoms is inhibited thus steering the selectivity. The presence of subsurface oxygen atoms is also concluded from a joint work with experimental and theoretical efforts of X-ray photoelectron spectroscopy. The precursor region of CO desorption from Ru(0001) is studied with the transition potential method. In contrast, for the simulation of the X-ray spectroscopy results on p4g C/Ni(100), which is a surface reconstruction when carbon atoms adsorb on Ni(100), vibrational effects are also needed for understanding the experimental data.