The chemistry of actinides in aqueous solution is important, and it is essential to build adequate conceptual models and develop methods applicable for actinide systems. The complex electronic structure makes benchmarking necessary. In the thesis a prototype reaction of the water exchange reaction for uranyl(VI), for both ground and luminescent states, described with a six-water model, was used to study the applicability of density functional methods on actinides and different solvation models. An excellent agreement between the wave function methods CCSD(T) and MP2 was obtained in the ground state, implying that near-minimal CASPT2 can be used with confidence for the reaction in the luminescent state of uranyl(VI), while density functionals are not suited to describe energetics for this type of reaction. There was an ambiguity concerning the position of the waters in the second hydration sphere. This issue was resolved by investigating a larger model, and prop- erly used the six-water model was found to adequately describe the water exchange reaction. The effect of solvation was investigated by comparing the results from conductor-like polarizable continuum models using two cavity models. Scattered numbers made it difficult to determine which solvation model to use. The final conclusion was that the water exchange reaction in the luminescent state of uranyl(VI) should be addressed with near-minimal CASPT2 and a solvation model without explicit cavities for hydrogens. Finally it was shown that no new chemistry appears in the luminescent state for this reaction. The thesis includes a methodological investigation of a multi-reference density functional method based on a range separation of the two-electron interaction. The method depends on a universal parameter, which has been determined for lighter elements. It is shown here that the same parameter could be used for actinides, a prerequisite for further development of the method. The results are in that sense promising.

The folding mechanism of several β-structures (e.g., β-hairpins and β-sheets) was studied using newly developed enhanced sampling methods along with MD simulations in all implicit solvent environments. The influence of different implicit solvent models on the folding simulation of β-structure was also tested. Through the analysis of the free energy landscape as the function of several suitable reaction coordinates, we observed that the folding of β-hairpins is actually a two-state transition. In addition, the folding free energy landscapes for those related hairpins indicate the apparent sequence dependence, which demonstrates different folding mechanisms of similar β-structures of varied sequence. We also found that the stability of backbone hydrogen bonds is determined by the turn sequence and the composition of hydrophobic core cluster in β-structures. Neither of these findings was reported before. The processive movement of kinesin was also studied at the mesoscopic level. We developed a simple physical model to understand the asymmetric hand-over-hand mechanism of the kinesin walking on the microtubule. The hand-over-hand motion of the conventional kinesin is reproduced in our model and good agreement is achieved between calculated and experimental results. The experimentally observed limping of the truncated kinesin is also perfectly described by our model. The global conformational change of kinesin heads (e.g., the power stroke of neck-linkers which works as lever-arms during the kinesin walking, the transition between open and closed states of the switch region of the nucleotide binding domain in each head induced by the nucleotide binding and release) was studied for both dimeric and monomeric kinesins using a coarse-grained model, anisotropic network model (ANM). At the same time Langevin mode analysis was used to study the solvent influence on the motions of the kinesin head mimicked by ANM. Additionally, the correlation between the neck-linker and the nucleotide binding site was also studied for dimeric and monomeric kinesins. The former shows the apparent correlation between two subdomains whereas the latter does not, which may explain the experimental observation that only the dimeric kinesin is capable of walking processively on the microtubule.

This dissertation describes the implementation of the molecular electronic structure calculations with an implicit solvent model using coupled-cluster theory. The theory for and the implementation of the solvent reaction field method (SCRF) and the reference interaction site model (RISM) at the coupled-cluster singles and doubles (CCSD) are presented. In the SCRF model a solute molecule is placed in a spherical cavity, and the outer solvent is represented by a dielectric continuum, which is characterized by the dielectric constant of the solvent. The reaction field is introduced to the system by using the multipole moment expansion of the electronic structure of the solute molecule and the dielectric constant. The SCRF method has been used to calculate the conformational equilibrium and the rotational barriers of 1,2-dichloroethane in vacuum and in different solvents. The calculated results are compared with experimental values. In addition, the solvent effects on the energetics of the mechanism of nitration of benzene are reported using the implemented CCSD-SCRF model. The idea of RISM is to replace the reaction field in continuum models by a microscopic expression in terms of the site-site radial distribution functions between solute and solvent, which can be calculated from the RISM integral equations. The statistical solvent distribution around the solute is determined based on the electronic structure of the solute, while the electronic structure of solute is influenced by the surrounding solvent distribution. Therefore, the wave function and the RISM equations are solved self-consistently with CCSD. Pair correlation functions, partial atomic charges, and solvation free energies of water and N-methylacetamide are calculated in liquid water using proposed theory. Both the CC-SCRF and CC-RISM methods have been implemented in a developmental version of the Q-Chem 3.2 quantum chemistry package.