Near-Perfect Absorption of Light by Coherent Plasmon-Exciton States

Strong Coupling of Two-Dimensional Excitons and Plasmonic Photonic Crystals: Microscopic Theory Reveals Triplet Spectra

Monolayers of transition metal dichalcogenides (TMDCs) are direct-gap semiconductors with strong light-matter interactions featuring tightly bound excitons, while plasmonic crystals (PCs), consisting of metal nanoparticles that act as meta-atoms, exhibit collective plasmon modes and allow one to tailor electric fields on the nanoscale. Recent experiments show that TMDC-PC hybrids can reach the strong-coupling limit between excitons and plasmons, forming new quasiparticles, so-called plexcitons. To describe this coupling theoretically, we develop a self-consistent Maxwell-Bloch theory for TMDC-PC hybrid structures, which allows us to compute the scattered light in the near- and far-fields explicitly and provide guidance for experimental studies. One of the key findings of the developed theory is the necessity to differentiate between bright and originally momentum-dark excitons. Our calculations reveal a spectral splitting signature of strong coupling of more than 100 meV in gold-MoSe2 structures with 30 nm nanoparticles, manifesting in a hybridization of the plasmon mode with momentum-dark excitons into two effective plexcitonic bands. The semianalytical theory allows us to directly infer the characteristic asymmetric line shape of the hybrid spectra in the strong coupling regime from the energy distribution of the momentum-dark excitons. In addition to the hybridized states, we find a remaining excitonic mode with significantly smaller coupling to the plasmonic near-field, emitting directly into the far-field. Thus, hybrid spectra in the strong coupling regime can contain three emission peaks.

Quantum-Coherence-Enhanced Hot-Electron Injection under Modal Strong Coupling

Modal strong coupling between localized surface plasmon resonance and a Fabry-Pérot nanocavity has been studied to improve the quantum efficiency of artificial photosynthesis. In this research, we employed Au nanodisk/titanium dioxide/Au film modal strong coupling structures to investigate the mechanism of quantum efficiency enhancement. We found that the quantum coherence within the structures enhances the apparent quantum efficiency of the hot-electron injection from the Au nanodisks to the titanium dioxide layer. Under near-field mapping using photoemission electron microscopy, the existence of quantum coherence was directly observed. Furthermore, the coherence area was quantitatively evaluated by analyzing the relationship between the splitting energy and the particle number density of the Au nanodisks. This quantum-coherence-enhanced hot-electron injection is supported by our theoretical model. Based on these results, applying quantum coherence to photochemical reaction systems is expected to effectively enhance reaction efficiencies.

Optimal Geometry for Plasmonic Hot-Carrier Extraction in Metal-Semiconductor Nanocrystals

Plasmon-induced energy and charge transfer from metal nanostructures hold great potential for harvesting solar energy. Presently, the efficiencies of charge-carrier extraction are still low due to the competitive ultrafast mechanisms of plasmon relaxation. Using single-particle electron energy loss spectroscopy, we correlate the geometrical and compositional details of individual nanostructures to their carrier extraction efficiencies. By removing ensemble effects, we are able to show a direct structure-function relationship that permits the rational design of the most efficient metal-semiconductor nanostructures for energy harvesting applications. In particular, by developing a hybrid system comprising Au nanorods with epitaxially grown CdSe tips, we are able to control and enhance charge extraction. We show that optimal structures can have efficiencies as high as 45%. The quality of the Au-CdSe interface and the dimensions of the Au rod and CdSe tip are shown to be critical for achieving these high efficiencies of chemical interface damping.