Time-dependent Nonadiabatic Investigations on the Photo-generated Carrier Dynamics at TiO2/molecule Interfaces

Band gap excitation of TiO2 instigates photocatalytic reactions. In the presence of molecular adsorbates the generated holes and electrons act as the reactants in the oxidation and reduction processes for solar fuel generation. In order to understand photocatalytic processes through theoretical methods one must go beyond theoretical descriptions of the static molecular end electronic structure, because the nuclear and electronic interactions leading to chemistry are strongly coupled. We propose to carry out ab initio theoretical investigations on dynamics of photogenerated holes and electrons based on time-dependent nonadiabatic molecular dynamics (NAMD) method. We will focus on the H2O/TiO2, CH3OH/TiO2 and CO2/TiO2 chemisorption systems, which are tractable by theory and for which extensive chemical, STM, and time-resolved two photon photoemission (TR-2PP) studies exist. We plan to study the interfacial transfer and decay dynamics of photogenerated carriers. The coupling with phonons and proton transfer will be also investigated. The proposed theoretical research will couple strongly with the ongoing TR-2PP spectroscopy experiments at the University of Pittsburgh.
The research on photocatalytic processes at metal oxide surfaces is directed to the clean and sustainable energy generation, environmental remediation, and climate change. The research addresses CESD goals to 1) strengthen model representations of energy and other human systems such as dynamic emissions; 2) advance understanding and process representation of the couplings involving energy, carbon, and water cycles; and 3) to use of EMSL high-performance computing to characterization and understanding of the experimental and theoretical basis of molecular-level reactions that occur in energy transduction and carbon recapture systems. The results of the proposed research are likely to have an impact on the fundamental molecular science for reducing the human impact on global climate and environment.
For the proposed NAMD simulations, usually we need to perform an ab initio molecular dynamics calculation up to 10 ps, which means 10,000 steps of electronic structure optimization. And for different structures and initial conditions, we need to perform many such calculations. Therefore, we believe the supercomputing facilities at EMSL are essential for our large scale computing needs. Neither PI has in-house computing facilities that could accomplish the proposed research. We request 150,000 node-hrs computing time on the Cascade supercomputer for the planned theoretical studies.