Open Access Version

Abstract:
The study of structure-property relationships is one of the key steps in developing efficient materials for challenging processes of our time, e.g. energy conversion and charge-transfer processes. In this regard, especially computer simulation techniques that enable insights on an atomistic scale are a powerful tool for finding new ways of improving material properties. However, since such predictions become increasingly difficult to perform depending on factors like energy gauges, extended system sizes and timescales of the processes, highly advanced simulation techniques and carefully selected model systems are required. In this dissertation, three examples for studying material properties on different timescales (nanoseconds to femtoseconds), energetic regimes (infra-red to X-ray response) and system sizes (nanoscopic to molecular) are presented for applications in energy and charge-transfer-related technologies. In the first study, the aim is to find the best-possible arrangement for a pair of charged quantum dots that drive the interatomic Coulombic decay (ICD) process after excitation with infra-red light. In this process, one of the charged quantum dots is excited by an external field to drive an electron ejection from a neighboring quantum dot. For describing such a highly correlated system of several nanometers in size including electron transitions into the continuum, the effective mass approximation (EMA) is applied and the very flexible multi-configurational time-dependent Hartree (MCTDH) electron dynamics description is utilized. Upon quantum dot size variations, it is found that there exists an intricate balance between the polarization of the electron cloud within the quantum dots and the Coulomb repulsion between them, which drives the process. This balance leads to general curves of size ratios that lead to maximum ICD rates for a given distance between the quantum dots. In a second project, the capabilities and limitations of a database approach are explored for interpreting X-ray absorption (XA) responses of amorphous substances. The approach is based on the fact that the absorption spectrum of any material can be obtained as a sum of all individual atom's responses in their exact environment. By establishing a database of XA fingerprint spectra for carbon atoms in unique surroundings up to a fixed radius, the most prominent oxidized functional groups are identified in different experimental samples of graphene oxides. The database of XA spectra is obtained by applying time-dependent density functional theory (TDDFT) calculations of local excitations in arbitrarily functionalized graphene oxide model molecules. The non-local effects following such high-energy excitations are shown to be in principle recoverable by expanding the radius for distinguishing the unique surroundings in this database. Conclusively, the method is suitable for extracting local structural patterns from amorphous materials. At last, the first partial reaction of the photochemical water-splitting on the edge of nitrogen-doped graphene oxides (NGO) is investigated. In this proton-coupled electron transfer (PCET), one electron and one proton of the water molecule are transferred onto the NGO catalyst in a concerted way. By performing TDDFT calculations on a large set of model molecules, the structural characteristics that lead to an efficient charge transfer are identified. Therefore, time-dependent configuration interaction (TDCI) calculations are applied for the electron dynamics that include explicit non-adiabatic coupling terms describing the hydrogen transfer. The results show that structures with conjugated ketone groups in the vicinity of the proton-accepting nitrogen atom are especially well suited for the electron transfer. Furthermore, the rate of the overall PCET process is strongly limited by the initial electronic excitation dynamics.