Bloino, Julien (2008) Development and application of time dependent and time independent models for the study of spectroscopic properties in compounds of biological interest. [Tesi di dottorato] (Unpublished)

Preview

PDF Bloino_Julien_these.pdf Download (3MB) | Preview

Collection description

In this thesis, we propose a general and effective approach to compute vibrationally-resolved electronic spectra from first principles. This method is integrated in a versatile quantum chemical computational package and offers a complete "in silico" procedure starting from the geometry optimization to the generation of the spectrum. The theoretical background and methods to evaluate the overlap integrals are presented, along with a discussion of strategies for an efficient evaluation of spectra of large systems, which features a huge number of possible vibronic transitions. The presented procedure relies on the general-purpose method to select "a priori" the transitions that should be calculated by estimating their probability. The implemented method uses a partition of the transitions by groups called "classes", which permits the usage of several computational schemes to speed up the calculations. The details of the procedure and the possibilities of fine-tuning of the calculations are presented, as well as an insight into its internal workout. The integrated approach to compute vibrationally resolved optical spectra can be applied to a large variety of systems ranging from small molecules in the gas phase to macrosystems in condensed phases, whenever nonadiabatic couplings are negligible and the harmonic approximation is reliable. The given examples of absorption spectrum of S1 <- S0 electronic transitions of anisole, photodetachment spectrum of SF6-, emission T1 -> S0 phosphorescence spectrum of chlorophyll \c2, UV spectrum of acrolein in the gas phase and aqueous solution, a photoelectron spectrum of adenine adsorbed on the Si(100) surface, and porphyrin are chosen to illustrate the possibilities of the procedure and some of its characteristics. It is shown that despite the fact that our computational scheme has been tailored for large systems, it can be utilized as well to generate high quality spectra for small systems. Moreover, good quality spectra can be effectively computed even for large systems with hundreds of normal modes, paving the route to spectroscopic studies of systems of direct biological and/or technological interest.

Downloads

Actions (login required)

View Item