Schneider, William F./Wolverton, Chris U. Notre Dame/Northwestern University
Under real world conditions the surface of a heterogeneous metal catalyst is covered with adsorbates, and the interactions of these adsorbates with each other and collectively with the catalyst surface can have a major impact on catalytic function. Atomic-level simulations of reactions at metal surfaces often can account for these interactions in only approximate ways, because of the large number (e.g. nearest-neighbor, next-nearest-neighbor, etc.) possible in even the simplest system. In this project, cluster expansion (CE) techniques developed in the alloy theory community to treat bulk ordering problems are extended to the description of adsorbate interactions at metal and bimetallic alloy surfaces. These CEs are parameterized against accurate density functional theory (DFT) calculations of the relative energies of a small number of adsorbate arrangements. Once parameterized, they provide a predictive model for the energetics of any arrangement of atoms at the surface-information that is used in Monte Carlo simulations to predict preferred adsorption geometries, surface orderings, surface phase diagrams, and adsorption thermodynamics. This same approach is similarly used to capture the effects of local adsorbate order on the activation energies of surface reactions.
The few extant examples of this surface DFT-CE approach are limited to single adsorbates. This project extends the approach in two ways key to their wider application in heterogeneous catalysis: to ternary" surfaces, which can capture the behavior of multiple adsorbates at a surface, and to CEs that capture the coupling between the surface composition of a bimetallic catalyst and adsorption of reactants. These tools are applied to the problem of O2 activation and catalytic oxidation at Pt and Au metal surfaces-relevant to problems from low-temperature fuel cells to environmental catalysis. CEs are constructed and used to predict the thermodynamics (through equilibrium Monte Carlo) and kinetics (through kinetic Monte Carlo) of dissociative oxygen adsorption and to the reaction of oxygen with an NO reductant as a function of external reactant conditions (T, PO2 ...). To evaluate the potential for tuning catalytic activity by alloying, CEs are also constructed and used to describe the equilibrium surface composition and oxygen reactivity of Pt-Au and Pt-Ag alloys in a number of interaction scenarios.
This work probes the limits of the DFT-CE methodology for describing surface reactivity and provides high-quality benchmarks for more approximate treatments.
The work done here crosses disciplinary boundaries and is built around the complementary expertise of the two PIs in surface reactivity (Schneider, Notre Dame) and alloy theory (Wolverton, Northwestern). By bridging these two communities in a highly collaborative effort, this work brings new understanding, new capabilities, and a strong potential for unanticipated innovation in heterogeneous catalysis. Broad dissemination of results to the catalysis and materials science communities further promotes cross-fertilization of ideas. In addition, the specific problem of interest-the catalytic activation of O2-is of utmost fundamental and practical importance to society.
The program provides a unique training environment for several chemical engineering and materials science graduate students in applied simulation and interdisciplinary research, facilitated by the close proximity of the partner institutions. The PIs both maintain and promote diverse research environments by involving underrepresented groups and engaging undergraduates in research. The work also supports the development of regular and summer short course curricula in simulation practice and application in both groups.
In summary, this program synthesizes education and training within a collaborative, interdisciplinary, multi-institutional research program developing new simulation tools and modeling approaches in the context of topical problems in heterogeneous catalysis.
