(32a) First-Principles Multiscale Modeling of CO Oxidation on Polycrystalline RuO2 in a Fixed Bed Reactor

CO oxidation is a widely employed prototype reaction for studying reactivity differences between different metal and metal oxide catalysts, and improved methods for simulating catalytic processes. Ru-based catalysts have been identified as being particularly active for CO oxidation at atmospheric pressures, despite their low activity under ultra-high vacuum conditions compared to similar catalysts.[1] There is still disagreement about the active phase under working conditions, despite significant experimental effort[2-4] Theoretical methods, in particular density functional theory (DFT) and kinetic Monte Carlo (KMC), have also been employed in an effort to elucidate the active phase under working conditions.[5, 6] Such efforts are hampered by the fact that neither DFT nor KMC incorporate explicit reactor-level effects that may induce emergent behavior, and reactor-level models incorporating KMC models are rare due to the challenges of coupling KMC simulations with reactor-level models.[7]

In this work we demonstrate a method for coupling multiple DFT-based KMC models to a computation fluid dynamics (CFD) model of a fixed bed reactor. We then use this multiscale model to investigate the effect of operating conditions and catalyst composition on the rate of CO oxidation, and we compare our simulation results to steady state experimental data. We find significant differences between the model and experimental apparent activation energies (~100 kJ/mol). Reaction orders are in better agreement. Our model also shows that the apical (111) facet of RuO₂ is several orders of magnitude more active than the (110) lateral facet of RuO₂ at the investigated conditions, in line with experimental evidence.[4] Overall, the nature of the active phase for Ru/RuO₂ catalysts under working conditions is still an open question, and more work must be done to develop improved kinetic models. The method used in this work provides one thrust to be followed for better understanding of Co oxidation over Ru/RuO2, as well as for other catalytic reactions where coupling of DFT calculated transition states to reactor-scale simulations is desirable.

1. Peden, C.H.F. and D.W. Goodman, Kinetics of carbon monoxide oxidation over ruthenium(0001). The Journal of Physical Chemistry, 1986. 90(7): p. 1360-1365.

2. Gao, F., et al., CO oxidation over Ru(0001) at near-atmospheric pressures: From chemisorbed oxygen to RuO2. Surface Science, 2009. 603(8): p. 1126-1134.

3. Assmann, J., et al., Heterogeneous oxidation catalysis on ruthenium: bridging the pressure and materials gaps and beyond. Journal of Physics-Condensed Matter, 2008. 20(18).

4. Rosenthal, D., et al., On the CO-Oxidation over Oxygenated Ruthenium. Zeitschrift Fur Physikalische Chemie-International Journal of Research in Physical Chemistry & Chemical Physics, 2009. 223(1-2): p. 183-207.

5. Reuter, K. and M. Scheffler, First-principles kinetic Monte Carlo simulations for heterogeneous catalysis: Application to the CO oxidation at RuO2(110). Physical Review B, 2006. 73(4).

6. Hess, F., et al., "First-Principles" kinetic monte carlo simulations revisited: CO oxidation over RuO2(110). Journal of Computational Chemistry, 2012. 33(7): p. 757-766.

7. Matera, S., et al., Predictive-Quality Surface Reaction Chemistry in Real Reactor Models: Integrating First-Principles Kinetic Monte Carlo Simulations into Computational Fluid Dynamics. Acs Catalysis, 2014. 4(11): p. 4081-4092.