Transient absorption spectroscopy is a powerful tool to monitor the out-of-equilibrium optical response of photoexcited semiconductors. When this method is applied to two-dimensional semiconductors deposited on different substrates, the excited state optical properties are inferred from the pump-induced changes in the transmission/reflection of the probe, i.e., ΔT/T or ΔR/R. Transient optical spectra are often interpreted as the manifestation of the intrinsic optical response of the monolayer, including effects such as the reduction of the exciton oscillator strength, electron-phonon coupling or many-body interactions like bandgap renormalization, trion or biexciton formation. Here we scrutinize the assumption that one can determine the non-equilibrium optical response of the TMD without accounting for the substrate used in the experiment. We systematically investigate the effect of the substrate on the broadband transient optical response of monolayer MoS2 (1L-MoS2) by measuring ΔT/T and ΔR/R with different excitation photon energies. Employing the boundary conditions given by the Fresnel equations, we analyze the transient transmission/reflection spectra across the main excitonic resonances of 1L-MoS2. We show that pure interference effects induced by the different substrates explain the substantial differences (i.e., intensity, peak energy and exciton linewidth) observed in the transient spectra of the same monolayer. We thus demonstrate that the substrate strongly affects the magnitude of the exciton energy shift and the change of the oscillator strength in the transient optical spectra. By highlighting the key role played by the substrate, our results set the stage for a unified interpretation of the transient response of optoelectronic devices based on a broad class of TMDs.
CC BY 4.0
