Extended Data Fig. 2: Steady state characterisation of BSWP.

(a) A sketch of the Kretschmann spectral imaging set-up used for either angle-resolved or spatially resolved reflection/emission steady-state measurements. (b,c) Dispersion of the BSWP modes measured via angle-resolved reflection (b) and emission (c) spectroscopy (represented by the false-colour plots). The sharp signal corresponds to the lower BSW polariton mode, while the white dashed lines represent the simulated dispersion (using the T-matrix method). The solid white lines indicate the bare BSW dispersion and the exciton energy (fixed at 2.13 eV), the red dashed line indicates the light line and the inset in (b) shows the typical linewidth of the reflection dip corresponding to the BSWP resonance, as measured around 1.91 eV. Note that in the reflection measurement (b), unlike similar angle-resolved reflectivity measurements conducted on polaritons in normal Fabry–Perot cavities,^7,49 here only the lower BSWP branch is observed. Since such reflectometry measurements are sensitive to absorption into the modes of the system the fact that the upper polariton is not observed probably results from inefficient coupling between incoming photons and the upper polariton modes via the prism. Moreover, in the emission measurements (c) the upper polariton is also missing, which is consistent with previous measurements and results from the fast, nonradiative decay of the upper polaritons^49,60.