Nite, Jacob M., authorRappé, Anthony K., advisorShores, Matthew, committee memberShi, Yian, committee memberLee, Siu Au, committee member2018-06-122019-06-072018https://hdl.handle.net/10217/189298Transition metal complexes have seen an increased use as photocatalysts for organic reactions in recent literature, mostly involving the Ru(II)(bpy)3 family of catalysts. Due to the rarity of ruthenium in the Earth's crust, alternative catalysts using Earth abundant materials are desirable. Recent literature has shown that chromium based catalysts show great promise as a replacement for ruthenium for some reactions. The mechanisms of these first-row transition metal complexes are significantly more complex than those of the second and third row. The excited state complexities of first-row transition metal complexes are challenges for both experimental and theoretical research. The complexities of the excited states require theoretical methods beyond the standard single reference methods commonly used in the literature. Through the use of recent multi-reference post Hartree Fock (HF) methods as well as a new multi-reference density functional theory (DFT), insights into the character of chromium-based photocatalysts were examined. A new multi-determinant DFT method named few-determinant density functional theory (FD-DFT) was described. FD-DFT incorporates multiple DFT determinants using a finite difference approach to calculate the exchanges between multiple determinants for open shell multiplets. The method is implemented in a generalized bond valence (GVB) wave function, and can be converged through an SCF procedure. The system was benchmarked using oxygen atom and diatomic oxygen as well as atomic systems with more open shell orbitals. The benchmarking shows stability across many different functional choices, and gives good excitation energies with and without SCF convergence. The Cr(III)(AcAc)3 system has been long studied for its unique excited state properties that defy the standard cascade model for excited state relaxation. The tris(1,3-propanedionato)chromium(iii) (Cr(III)(PDO)3) complex was studied as an analog to the Cr(III)(AcAc)3 system to understand the excited state pathway between the initial excited 4T2g state and the long lived 2Eg state. Using the FD-DFT method as well as the multi-reference spectroscopy oriented configuration interaction (SORCI) method, the initial excited state energies were studied compared to previous perturbation theory (PT) approaches. Both SORCI and FD-DFT calculate reasonable 2Eg excitation energies, an improvement over earlier results. The SORCI method was also used to map the potential energy curve between the initial 4T2g excited state and its fully relaxed distorted structure. The pathway agrees with previous experimental and theoretical studies showing that a transitionless path exists between the quartet and doublet states, but spin-orbit coupling calculations suggest that a direct path between the 4T2g and 2Eg is possible rather than needing a internal conversion step to the lowest 2Eg state. Chromium-based photocatalysts have been recently studied in the literature as having a competitive mechanism between the reaction substrate and O2 whereby the O2 quenches the excited catalyst. Using the combined Cr(III)(PDO)3 • O2 system, the likely states by which this quenching event occurs were studied with FD-DFT as well as recent multi-reference PT approaches. Comparing the excited state calculated using the multi-reference based methods to standards DFT calculations shows the inability of single-determinant methods to correctly produce the proper excited state character even when obtaining somewhat reasonable energies. The excited state responsible for the quenching of the excited complex is identified using spin density plots of the CASSCF calculations. The search for suitable first-row transition metals requires a search across possible ligands and metal centers. Using the success of chromium-based catalysts, isoelectronic vanadium catalysts were studied to identify any potential differences between the complexes as well as identify the utility of vanadium-based catalysts. Using a variety of methods, including TDDFT-based absorption spectra, vibrational component plots of the excited state distortions, and SORCI potential energy curves (PEC), the differences between the chromium and vanadium catalysts were examined. It was found that vanadium catalysts absorptions are shifted significantly from chromium complexes and the vanadium excited states disperse the unpaired electron over the complex instead of localizing it on the metal center. The distortions in the chromium-based catalysts have a greater amount of asymmetric vibrational character compared to vanadium, which shows mostly symmetric behavior. Lastly, the SORCI PECs show that, unlike chromium, the doublet curves do not intersect the quartet curves, making a transition to a long lived doublet state a significantly slower process. The results highlight significant differences between the complexes even with ligand structure is controlled.born digitaldoctoral dissertationsengCopyright and other restrictions may apply. User is responsible for compliance with all applicable laws. For information about copyright law, please see https://libguides.colostate.edu/copyright.photocatalysisDFTText