Dark and Bright Excitons in Halide Perovskite Nanoplatelets

TL;DR

This study investigates the exciton fine structure in halide perovskite nanoplatelets, revealing thickness-dependent bright-dark exciton splitting and polarization effects, which influence their optical properties and potential for fast luminescence.

Contribution

The paper introduces a combined experimental and modeling approach to elucidate the exciton fine structure in perovskite nanoplatelets, highlighting the impact of thickness and polarization on exciton states.

Findings

Bright-dark exciton splitting reaches 32.3 meV in 2ML NPLs.

Bright exciton states show polarization-dependent splitting of 5-16 meV.

Dark excitons significantly influence optical properties even at room temperature.

Abstract

Semiconductor nanoplatelets (NPLs), with their large exciton binding energy, narrow photoluminescence (PL), and absence of dielectric screening for photons emitted normal to the NPL surface, could be expected to become the fastest luminophores amongst all colloidal nanostructures. However, super-fast emission is suppressed by a dark (optically passive) exciton ground state, substantially split from a higher-lying bright (optically active) state. Here, the exciton fine structure in 2-8 monolayer (ML) thick Cs_{n-1}Pb_nBr_{3n+1} NPLs is revealed by merging temperature-resolved PL spectra and time-resolved PL decay with an effective mass modeling taking quantum confinement and dielectric confinement anisotropy into account. This approach exposes a thickness-dependent bright-dark exciton splitting reaching 32.3meV for the 2ML NPLs. The model also reveals a 5-16 meV splitting of the bright exciton states with transition dipoles polarized parallel and perpendicular to the NPL surfaces, the order of which is reversed for the thinnest NPLs, as confirmed by TR-PL measurements. Accordingly, the individual bright states must be taken into account, while the dark exciton state strongly affects the optical properties of the thinnest NPLs even at room temperature. Significantly, the derived model can be generalized for any isotropically or anisotropically confined nanostructure.