Computational Studies of Electronic Excited States, or: How I Learned to Stop Worrying and Love Multireference Methods

• Author / Creator

  Oakley, Meagan

• Modelling electronic excited states is important in the search to understand molecular properties. Vacuum ultraviolet (VUV) spectroscopy can be used to identify different isomers in complicated mixtures of many molecules. In this Thesis, calculated VUV spectra were compared with spectra of experimental mixtures to benchmark appropriate computational methods. Because the benchmark molecule, 1-bromo-1-propene, contains a heavy atom, both all-electron and model core potential basis sets were investigated. Time-dependent density functional theory (TD-DFT) can accurately compute electronic excited states at low-energy excitations and was cross-checked at higher energies against results from the symmetry adapted cluster-configuration interaction (SAC-CI) method. TD-DFT was determined to be satisfactory at low energies; however, excitation energies can deviate by 0.5 eV at high energies. TD-DFT with both all-electron and model core potential basis sets produced satisfactory excitation energies for the lower excited states. This method was also able to predict spectra produced experimentally, including a mixture of isomers (cis- and trans-1-bromo-1- propene), even if the oscillator strengths were underestimated.
  The potential energy surface for the thermal decomposition reaction P4 → 2P2 was computed along the C2v reaction trajectory. Single-reference methods were not suitable for describing this complex bond-breaking process, so two multiconfigurational methods, namely, multi-state complete active space second-order perturbation theory (MS-CASPT2) and multiconfiguration pair-density functional theory (MC- PDFT), were used with the aim of determining the accuracy and efficiency of these methods for this process. Several active spaces and basis sets were explored. It was found that the MC-PDFT method was up to 900 times faster than MS-CASPT2 while providing similar accuracy.
  A new method, ∆DFT/MIX, was proposed and calibrated for use in calculating core electron binding energies. Chemically relevant test sets were used to determine the most accurate functionals out of the 70 density functionals included in GAMESS. The best three functionals, B3LYP, TPSSm, and BLYP, were used to calculate the 1s electron binding energies of nucleic acid base tautomers, and the results were com- pared to experimental values to demonstrate accuracy and sensitivity of the method. Previously suggested methods such as ADC(4) and ∆MP2/MIX are as accurate as the new method; the overall mean absolute deviation of ∆DFT/MIX is 0.19 eV. Comparing calculation time shows that using DFT instead of MP2 is much less computationally costly for larger molecules.
  Lastly, a study in designing chemosensors for the detection of heavy metals was done using numerous crown structures for preferential binding to Hg2+. Binding affinity was calculated with respect to other hydrated metal ions in order to incorporate explicit solvent effects (water and acetonitrile). The ionophore binding affinity was selective to highly charged metal ions (3+, 4+), but a narrowed metal ion study (Zn2+, Cd2+, Hg2+) and the full chemosensor structure showed that binding to Hg2+ is preferential. Electronic excited states of the full chemosensor were calculated to determine the metal complex with a bright state with the largest difference compared to the excited states of the parent ligand. Ultimately, the chemosensor with the best binding affinity and the largest difference in excitation energy (∆∆E) was 18C-O4S2-meta-a, with a binding affinity for Hg2+ of -67 kJ/mol with respect to Zn2+, and -56 kJ/mol with respect to Cd2+ and a ∆∆E of 1.4 eV compared to the parent ligand.

• Subjects / Keywords

  • Electronic Excited States
  • Computational Chemistry

• Graduation date

  Fall 2019

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/r3-rgat-re47

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.