Predictive Theory and Modelling of Heterogeneous Interfaces

While much progress has been made in the microscopic characterization of bulk materials, understanding the behavior, at the atomic scale, of complex materials interfaces remains a challenging problem in condensed matter physics and materials science. Development of predictive theory and modeling for interfacial systems is not only important to aid experimental interpretations but is also crucial to accelerate materials design efforts for a broad range of technological applications from the semiconductor industry to the development of new sources of energy.;This dissertation presents a computational framework that combines first-principles molecular dynamics simulations for the determination of interfacial atomic structures, and advanced electronic structure methods for the description of electronic properties to understand, predict and design materials interfaces. In particular, the research presented in this thesis was carried out in two parallel directions: 1. Predictions of the interfacial atomic structure of complex materials interfaces using first-principles molecular dynamics, and validation of the predictions by relating structural models to experiments, e.g., data from X-ray, infrared spectra and sum-frequency generation spectroscopy experiments. In particular, I investigated structural and dynamical properties of the ice Ih surface and of the Al2O 3/water interface. While the study of the ice surface is the first step towards the understanding of complex semiconductor/water interfaces, the Al 2O3/water interface represents a suitable prototype interface for extensive comparisons between theory and experiments. 2. Development of advanced first-principles techniques to study electronic states at interfaces and application of these techniques to gain insights on the relationship between between local interface structure and electronic properties. I devised a new technique to study excited states and photoemission spectra based on many-body perturbation theory, within the so-called GW approximation, which improves both the computational efficiency and accuracy of existing methodologies, and that can be employed to study realistic systems. In addition, in the thesis I employed the new GW technique to investigate a variety of systems including molecules, nanostructures, semiconducting interfaces, liquid water and simple aqueous solutions. These studies are crucial to build a fundamental understanding of the electronic structure of semiconductor/liquid water interfaces and of, e.g., water with dissolved ions under different pH conditions, interfaced with a photoelectrode.;Information provided by these two parallel research directions helped establish a structure-electronic properties-chemical reactivity paradigm, that is general and applicable to a large class of materials. An example presented in the thesis is that of functionalized Si surfaces interfaced with liquid water, in which I studied the effect of surface functionalization on the alignment between Si band edges and water redox potentials, and I suggested a possible approach to engineer and design semiconductor surfaces for photoelectrochemical water splitting.