Olivia Hull, Kansas State University
Harnessing the catalytic activity of plasmonic nanostructures recently has become a point of intense experimental interest. Plasmonic nanostructures have been shown to selectively drive reactions under mild conditions on molecules that are traditionally difficult to activate. Plasmon resonance – the collective sloshing of electrons upon irradiation with light of a particular frequency – is thought to play an important role in the activation process. However, a detailed mechanistic understanding of plasmon-mediated photocatalysis (PMP) is lacking. Computational methods presently cannot both maintain computational tractability and properly account for the complex, multiconfigurational electron dynamics of the plasmon mode. Herein, we present a theoretical framework that aims to remedy this issue. We have developed the theory to build a configuration interaction-type approximation into the non-adiabatic fewest-switches surface hopping (NA-FSSH) program PYXAID. Quantum mechanically, the plasmon mode can be described as a coherent addition of multiple single-particle excitations. Which single-particle excitations participate and the extent of their contribution can be obtained through linear-response time-dependent density functional theory (LR-TDDFT)-type calculations. Meanwhile, NA-FSSH computes electron dynamics by determining a "hopping" probability between electronic surfaces at each time step. The PYXAID program approximates these electronic surfaces as deriving from single electronic excitations and therefore cannot model plasmonic excitations. However, by determining which single-particle states participate in the plasmon mode via LR-TDDFT, generating a coefficient matrix whose elements are the LR-TDDFT calculated "weights" of the single-particle excitations, and performing a straightforward basis set change via matrix similarity transformation on the PYXAID-computed non-adiabatic coupling matrix, we can obtain NA-FSSH dynamics for plasmonic systems. This transformation allows the user to obtain time-dependent electron dynamics information on plasmonic excitations without adding significant computational overhead. This advance will enable the computational study of PMP within the NA-FSSH framework and thus provide deeper insights into its possible mechanisms.
Abstract Author(s): Olivia A. Hull, Christine M. Aikens