Fig. 2: Strong light–matter interaction regime and polariton condensation.

a The top panel shows the emission as a function of excitation fluence. This light-in light-out characteristic exhibits a nonlinear increase (power-law fit in green) above the threshold (P_th ~ 39 µJ cm^-2, dashed line), after a linear emission (power-law fit in magenta), and saturation at higher fluences. The bottom panel displays the center energy of the emission peak (blue squares) and the width (orange triangles) as a function of excitation fluence, showing a sudden narrowing at the threshold and a continuous blue-shift above. b Emission spectra below threshold (gray data points) and above threshold (black data points), and their fitted spectra (magenta and green, respectively). c Real-space emission images of the cavity when pumped below and d, above threshold at the same fluences as the spectra in (b) were obtained. e Reflectivity (gray scale) obtained from a two-dimensional (2D) rigorous coupled wave analysis (RCWA) calculation as a function of cavity length in the weak coupling regime with refractive index n = 1.86 = const. and k = 0; the different longitudinal modes appear as dark stripes in the gray scale data. The experimentally measured energies (red circles) are extracted above threshold, where the nonlinear emission yields a sharper, more intense peak: these energies clearly are not matching the slope of the weak-coupling theory. The exciton energy is represented with a yellow dashed line at 2.714 eV. The right panel shows the absorption (green, Abs) and emission (blue, Em) spectra of a bare MeLPPP thin film without a cavity. f Simulated reflectivity (RCWA) as a function of cavity length for cavities in the strong-coupling regime (gray scale), where the full refractive index dispersion n = n(λ) and k = k(λ) is used, as obtained from ellipsometry of the polymer layer. The experimentally measured energies of the polariton condensates are overlaid as red circles. To account for the 2D nature of the simulation, which neglects the vertical guiding layer structure that results in a lower effective refractive index, the theoretical curves are shifted slightly (+0.07 µm in cavity length) to achieve a good match with the experiment. This offset dominates over the small energy blueshift (at most 5 meV) occurring above threshold, which is approximately rigid and thus negligible in comparison with the applied horizontal correction (~11 meV). The right panel shows the absorption (green) and emission (blue) spectra