(616e) Ab Initio Molecular Dynamics Spectra for Characterization of Hydrated Supported Metal Oxide Catalysts

Vibrational spectroscopy is an essential tool to determine the structure and active site of supported catalytic materials. However, due to these materials' vastly complex configurational space, it is virtually impossible to accurately gain clear insights into these systems based solely on well-established single-crystal experiments. To circumvent this challenge, we turn to computational spectra to provide a direct mapping between structure and spectra.

We compute Raman and power spectra for molybdenum oxide on γ-alumina via ab initio molecular dynamics (AIMD) and compare them to conventional 0 K density functional theory spectra. AIMD spectra can capture anharmonicities and features present at finite temperatures making them more amendable to comparison with experimental data. We find that the hydration conditions, often neglected for modeling simplicity, significantly affect spectral signatures. Additional analysis of the isolating hydroxyl group frequencies from power spectra suggests that the changes seen experimentally in infrared spectra can be related to surface reconstruction upon anchoring MoO_x species on alumina. We also highlight system conditions that lead to different predicted frequencies between AIMD and traditional methods, which should be taken into consideration when directly comparing experimental and computational spectra.