Abstract
Theoretical gas-surface models that describe adsorption over a broad range of adsorbate concentrations can provide qualitative insight into chemical phenomena, such as subsurface adsorption, surface reconstruction, and industrial heterogeneous catalysis. However, most atomistic, quantum-mechanical models of gas-surface adsorption are limited to low adsorbate coverage due to the large computational cost of models built using many surface atoms and adsorbates. To investigate adsorption in the subsurface of a crystalline solid with increasing coverage, we present a lattice-gas adsorption model that includes surface and subsurface sites of the solid and is fully parametrized using density functional theory. We apply the model to study the competition between the surface and subsurface adsorption of atomic oxygen on the Ag(111) surface. Oxygen population distributions calculated using the model in combination with Monte Carlo simulations show the onset of subsurface adsorption above a total coverage of 0.375 monolayer and a greater accumulation of oxygen in the second than in the first subsurface at total coverages between 0.5 and 1 monolayer. Our simulations also show that oxygen atoms do not percolate into the bulk region of silver for total coverages of up to 1 monolayer, indicating that oxygen adsorbed in the subsurface is distinct from oxygen absorbed in the bulk in this coverage range. Computations of core-electron binding energies and projected density of states for the equilibrium oxygen distribution at 0.5 monolayer reveal qualitative differences in the oxygen-silver bonding at the surface and subsurface, suggesting that oxygen adsorbed in the two regions could play distinct roles in the surface chemistry.
Original language | English |
---|---|
Pages (from-to) | 5343-5353 |
Number of pages | 11 |
Journal | Journal of Physical Chemistry C |
Volume | 126 |
Issue number | 11 |
DOIs | |
State | Published - Mar 24 2022 |
Externally published | Yes |
Funding
This research was supported by the National Science Foundation CAREER grant CHE-1753273. Coverage-dependent DFT calculations in this work used the Stampede cluster at the Texas Advanced Computing Center, the Bridges cluster at the Pittsburgh Supercomputing Center, and the Comet cluster at the San Diego Supercomputer Center through the startup allocation CHE180046 at the Extreme Science and Engineering Discovery Environment (XSEDE). XSEDE is supported by National Science Foundation grant number ACI-1548562. Computations of the adsorption model and MC simulations were performed on the computational resources at the Infrastructure for Scientific Applications and Advanced Computing (ISAAC) supported by the University of Tennessee. S.R. thanks Daniel Killelea for helpful discussions.