Methods and Strategies for the Ab Initio Design of Novel Mn Oxide-Based Water Splitting Photocatalyst Materials

Abstract

Photoelectrochemical cells (PECs) use sunlight to drive endoergic reactions such as carbon dioxide reduction to fuels or water-splitting for renewable hydrogen production. However, materials that combine both the efficiency and low cost needed to make solar-powered catalysis a practical reality have yet to be discovered. This thesis presents methods and new design strategies for developing novel, efficient, robust, and inexpensive photocatalysts based on transition metal oxides (TMOs). Quantum mechanics methodologies are developed and tested for their ability to predict the properties of known materials and then used to predict how altering the composition by alloying and doping with abundant elements affects optical, electronic, transport, and catalytic properties.

Modeling TMOs poses significant challenges to density functional theory (DFT), the workhorse of electronic structure calculations. We test various DFT approximations for their ability to predict the ground state properties of MnO. Additionally, a first principles scheme for determining the band edge placements with respect to vacuum is proposed and tested for several materials. Greenfs function (GW) and embedded correlated wavefunction (ECW) theories are used to study excited state properties. We compare several ECW approaches that use electrostatic or orbital-free-DFT-based embedding potentials for predicting excitations within MgO as a prototype. We find ECW theory with an electrostatic description of the background significantly improves predicted energetics over standard DFT or non-embedded CW methods.

The first material considered for photocatalysis is MnO, the bio-inspired solid state analogue of the photosystem II active site. GW theory with input from hybrid DFT and ab initio DFT+U capably predicts the photoemission/inverse photoemission (PE/IPE) band gap and dielectric properties. An ab initio value of U-J = 3.5 eV for Mn"<super>2+</super>" was determined using unrestricted Hartree-Fock theory on cluster-size-converged electrostatically embedded clusters. The lowest-lying excitations in MnO, studied using ECW theory, are found to be single Mn d-d ligand field excitations (~2.5 eV, ~10<"super>8</super>" s lifetime), followed by double d-d excitations (~5.2 eV, ~10<"super>6</super>" s lifetime), Mn 3d-4s excitations (~6.3 eV, ~10"<super>-3</super>" s lifetime), and higher-lying O 2p-Mn 3d ligand-to-metal charge-transfer (LMCT) excitations (~10.1 eV, ~10"<super>-4</super>" s lifetime). The longer-lived transitions should exhibit better electron-hole pair separation and enhance photoconductivity depending on ease of carrier transport.

While MnO possesses suitable band edge energies, its band gap is too large for efficient sunlight absorption. We predict alloying MnO with ZnO in varying amounts reduces the PE/IPE band gap (to 2.6 eV for the 1:1 alloy) while preserving potential redox reactivity. Optical excitation studies show alloying lowers the LMCT transition to ~8.3 eV leaving all other absorption properties relatively unchanged. We find near degeneracies among spin-allowed and spin-forbidden LMCT states that could facilitate intersystem crossing (ISC) resulting in longer lifetimes. We suggest seeking other materials that exhibit similar LMCT excitations but that are visible-light activated as a design strategy for further enhancing photon conversion efficiencies. Additionally, several dopants (Al, Ga, In, Sc, Y, Ti, Sb, Gd, F (n-type dopants) and Li (a p-type dopant)) were assessed for their ability to enhance conductivity in MnO:ZnO. We find Ga, Sc, Ti, F, and Sb dopants create deep traps whereas In forms shallower traps that merit further investigation. In contrast, Y, Al, Gd, and Li dopants should increase the carrier concentration while maintaining favorable electron and hole transport pathways.

The adsorption and oxidation of water on MnO:ZnO(001) surface was studied with ab initio DFT+U calculations. The computed phase diagram for the water/MnO:ZnO(001) interface reveals the surface is quite hydrophilic with the half-dissociated 1 ML (2 ML) structure being most stable under water-poor (water-rich) conditions. For the gas phase water oxidation reaction, we compute a thermodynamic overpotential of 0.82 V without yet modeling reaction kinetics or solvation. The overpotential mainly results from the *OOH intermediate being too weakly bound to the surface because of a loss of resonance stabilization in the adsorbate. We suggest judicious doping as a way to stabilize *OOH and potentially reduce the overpotential to just 0.05 V (for 0.5 ML reaction coverage).