DEVELOPMENT OF HIGH LEVEL AB INITIO METHODS TO DESCRIBE NONADIABATIC EVENTS AND APPLICATIONS TO THE EXCITED STATES OF SMALL BIOLOGICAL MOLECULES

Abstract

The development of quantum mechanics has historically allowed researchers to theoretically explore the fundamental physical properties of atoms and molecules. Although quantum mechanics has been around for almost a century, its use was largely limited by the computational complexity it demanded. In the past decade, computer technology has evolved to the point where it is possible to perform calculations on biologically relevant systems. This has allowed us to corroborate results obtained from experiment as well as predict and explain phenomena that experiment cannot. Unfortunately, the field as a whole has not progressed to the point where high level methods, such as Multi-Reference Configuration Interaction (MRCI), are applicable to large molecular systems. Thus, to effectively study these systems, compromises must be made. In this work, two different approaches are taken to study the photophysical properties of systems such as DNA. In the first approach, a model system is formulated and studied in lieu of the larger target system. The excited state dynamics of 8-oxoguanine (8-oG) and its anion are studied in order to assess the possibility of taking part in an electron transfer mechanism to repair a nearby cyclobutane pyrimidine dimer (CPD). It is found that barriers on the anion S1 excited state surface prohibits easy access to conical intersections with the ground state, causing the anion to have a much longer excited state lifetime than the neutral form. Although much insight can be gained by this method, it is not uncommon for crucial interactions to be lost through simplification. In this case, when 8-oG is placed in an adenine dinucleotide, the π stacking interaction allows it to form a long lived radical base pair, which may be fundamental to its role in CPD repair. Unfortunately, it is impossible to carry out the same excited state calculations for the 8-oG/adenine dinucleotide due to computational cost. For reasons such as these, we also implement and benchmark a new approach to carrying out high level configuration interaction calculations in which the MRCI is expanded in the basis of high multiplicity natural orbitals (HMNOs). Specifically, the HMNO approach is implemented by expanding the MRCI wavefunction in the basis of natural orbitals generated from a ground state high multiplicity Configuration Interaction Singles and Doubles (CISD) calculation. Excited state calculations both at and away from the Franck-Condon region were performed to benchmark the ability of the HMNO approach using CISD and MRCI to reproduce standard MRCI energies. The ability of the HMNOs to be truncated was also explored, yielding efficient truncation criteria and guidelines for choosing the best basis set. It is found that the MRCI/HMNO approach yields energies that are in excellent agreement with standard MRCI while only requiring a fraction of the computational effort, possibly allowing it to be applied to larger systems such as nucleotide dimers.