In this work a systematic method for determining correlated wavefunctions of extended systems in the ground state as well as in excited states is presented. It allows to fully exploit the power of quantum-chemical programs designed for correlation calculations of finite molecules. Using localized Hartree-Fock (HF) orbitals (both occupied and virtual ones), an effective Hamiltonian which can easily be transferred from finite to infinite systems is built up. Correlation corrections to the matrix elements of the effective Hamiltonian are derived from clusters using an incremental scheme. To treat the correlation effects, multireference configuration interaction (MRCI) calculations with singly and doubly excited configurations (SD) are performed. This way one is able to generate both valence and conduction bands where all correlation effects in the excited states as well as in the ground state of the system are taken into account. An appropriate size-extensivity correction to the MRCI(SD) correlation energies is developed which takes into account the open-shell character of the excited states. This approach is applicable to a wide range of polymers and crystals. In the present work trans-polyacetylene is chosen as a test system. The corresponding band structure is obtained with the correlation of all electrons in the system being included on a very high level of sophistication. The account of correlation effects leads to substantial shifts of the "center-of-mass" positions of the bands (valence bands are shifted upwards and conduction bands downwards) and a flattening of all bands compared to the corresponding HF band structure. The method reaches the quantum-chemical level of accuracy. Further an extention of the above approach to excitons (optical excitations) in crystals is developed which allows to use standard quantum-chemical methods to describe the electron-hole pairs and to finally obtain excitonic bands.

In this work a systematic method for determining correlated wavefunctions of extended systems in the ground state as well as in excited states is presented. It allows to fully exploit the power of quantum-chemical programs designed for correlation calculations of finite molecules. Using localized Hartree-Fock (HF) orbitals (both occupied and virtual ones), an effective Hamiltonian which can easily be transferred from finite to infinite systems is built up. Correlation corrections to the matrix elements of the effective Hamiltonian are derived from clusters using an incremental scheme. To treat the correlation effects, multireference configuration interaction (MRCI) calculations with singly and doubly excited configurations (SD) are performed. This way one is able to generate both valence and conduction bands where all correlation effects in the excited states as well as in the ground state of the system are taken into account. An appropriate size-extensivity correction to the MRCI(SD) correlation energies is developed which takes into account the open-shell character of the excited states. This approach is applicable to a wide range of polymers and crystals. In the present work trans-polyacetylene is chosen as a test system. The corresponding band structure is obtained with the correlation of all electrons in the system being included on a very high level of sophistication. The account of correlation effects leads to substantial shifts of the "center-of-mass" positions of the bands (valence bands are shifted upwards and conduction bands downwards) and a flattening of all bands compared to the corresponding HF band structure. The method reaches the quantum-chemical level of accuracy. Further an extention of the above approach to excitons (optical excitations) in crystals is developed which allows to use standard quantum-chemical methods to describe the electron-hole pairs and to finally obtain excitonic bands.

Im Folgenden werden Transferprozesse in Nanohybridsystemen theoretisch untersucht. Diese Hybridsysteme sind vielversprechende Kandidaten für neue optoelektronische Anwendungen und erfahren daher ein erhebliches Forschungsinteresse. Jedoch beschränken sich Arbeiten darüber hauptsächlich auf experimentelle Untersuchungen und kaum auf die dazugehörige theoretische Beschreibung. Bei den theoretischen Betrachtungen treten entscheidende Limitierungen auf. Es werden entweder Details auf der atomaren Ebene vernachlässigt oder Systemgrößen betrachtet, die wesentlich kleiner als im Experiment sind. Diese Thesis zeigt, wie die bestehenden Theorien verbessert werden können und erweitert die bisherigen Untersuchungen durch die Betrachtung von vier neuen hoch relevanten Nanohybridsystemen. Das erste System ist eine Nanostruktur, die aus einem Au-Kern und einer CdS-Schale besteht. Beim zweiten System wurde eine ZnO/Para-Sexiphenyl Nanogrenzfläche untersucht. Die zwei anderen Systeme beinhalten jeweils einen CdSe-Nanokristall, der entweder mit einem Pheophorbide-a-Molekül oder mit einem röhrenförmigen Farbstoffaggregat wechselwirkt. In allen Systemen ist der Anregungsenergie-Transfer ein entscheidender Transfermechanismus und steht im Fokus dieser Arbeit. Die betrachteten Hybridsysteme bestehen aus zehntausenden Atomen und machen daher eine individuelle Berechnung der einzelnen Subsysteme sowie deren gegenseitiger Wechselwirkung notwendig. Die Halbleiter-Nanostrukturen werden mit der Tight-Binding-Methode und der Methode der Konfigurationswechselwirkung beschrieben. Für das molekulare System wird die Dichtefunktionaltheorie verwendet. Die dazugehörigen Rechnungen wurden von T. Plehn ausgeführt. Das metallische Nanoteilchen wird durch quantisierte Plasmon-Moden beschrieben. Die verwendeten Theorien ermöglichen eine Berechnung von Anregungsenergietransfer in Nanohybridsystemen von bisher nicht gekannter Systemgröße und Detailgrad. / In the following, transfer phenomena in nanohybrid systems are investigated theoretically. Such hybrid systems are promising candidates for novel optoelectronic devices and have attracted considerable interest. Despite a vast amount of experimental studies, only a small number of theoretical investigations exist so far. Furthermore, most of the theoretical work shows substantial limitations by either neglecting the atomistic details of the structure or drastically reducing the system size far below the typical device extension. The present thesis shows how existing theories can be improved. This thesis also expands previous theoretical investigations by developing models for four new and highly relevant nanohybrid systems. The first system is a spherical nanostructure consisting of an Au core and a CdS shell. By contrast, the second system resembles a finite nanointerface built up by a ZnO nanocrystal and a para-sexiphenyl aggregate. For the last two systems, a CdSe nanocrystal couples either to a pheophorbide-a molecule or to a tubular dye aggregate. In all of these systems, excitation energy transfer is an essential transfer mechanism and is, therefore, in the focus of this work. The considered hybrid systems consist of tens of thousands of atoms and, consequently, require an individual modeling of the constituents and their mutual coupling. For each material class, suitable methods are applied. The modeling of semiconductor nanocrystals is done by the tight-binding method, combined with a configuration interaction scheme. For the simulation of the molecular systems, the density functional theory is applied. T. Plehn performed the corresponding calculations. For the metal nanoparticle, a model based on quantized plasmon modes is utilized. As a consequence of these theories, excitation energy transfer calculations in hybrid systems are possible with unprecedented system size and complexity.