Computational Discoveryof Design Principles for Plasmon-DrivenBond Activation on Alloy Antenna Reactors

Plasmonic “antenna reactor” alloys, consisting of a plasmonic material doped with a catalytically active metal, show great promise for efficient photocatalysis. However, while simple, intuitive, and approximate design principles such as the Sabatier principle have been developed for thermal and electrocatalysis, similar design principles for plasmonic catalysts remain elusive. Here, we develop these simple design principles by using real-time, time-dependent density functional theory to study small molecule activation (CH₄, CO₂, H₂O, and N₂) on a number of Cu-based antenna reactors and elucidate trends. We first show that this technique gives results consistent with experimental plasmonic catalysis studies. We then identify promising, previously untested antenna reactors for these molecules. Next, we find that, for a given molecule, bond activation correlates with the size of the charge oscillations between the nanoparticle and molecule as quantified by the standard deviation over the propagation time. Furthermore, the orbital overlap between the dopant and molecule also roughly correlates with the bond activation. For CH₄, N₂, and H₂O, a greater overlap leads to higher activation. For CO₂, the trend is reversed because a greater overlap leads to higher chemical activation upon adsorption, which inhibits photoactivation. Hence, the orbital overlap can be used as a computationally efficient and intuitively simple predictor of photoactivation for the initial screening of plasmonic catalysts.