Date of Graduation
Master of Science in Chemistry
Chemistry and Biochemistry
General and accurate computational methodologies are currently lacking for large chemical systems. This is primarily due to the computational expense required to perform calculations on systems with one hundred or more atoms. Calculated spectroscopic properties could aid in the process of elucidating structural features of large biologically relevant molecules if accurate and inexpensive methods are developed. Towards this end the first steps were taken to design a general methodology for predicting NMR chemical shifts of large nucleic acid systems. It was found that HF and semi-empirical methods were not sufficient for optimization of nucleobases, and therefore larger nucleotide or nucleic acid systems. It was also found that there is little difference in performance between DFT methods for prediction of NMR shifts of nucleobases as long as hydrogen bonding requirements are satisfied. To potentially reduce the computational expense of calculating Raman activities, a new and potentially inexpensive numerical method was developed. This method utilizes volume changes as a basis for approximating polarizability changes over the course of molecular vibrations. Raman intensities calculated using this methodology were compared to experimentally obtained Raman intensities by linear regression. While a positive correlation was found further refinement is needed.
DFT, DNA, Raman, NMR, electronic volume
Organic Chemistry | Physical Chemistry
© William R. Ehrhardt
Ehrhardt, William R., "NMR Spectroscopic Properties of Nucleotides, and a New Method of Numeric Calculation of Raman Intensities for Organic Molecules" (2018). MSU Graduate Theses. 3299.