| Photoluminescence from direct-bandgap semiconductor quantum wells after non-resonant excitation is predominantly observed at energetic position of the 1s exciton resonance. The time evolution of the photoluminescence is generally interpreted as direct monitor of an excitonic population; a rise of the signal is interpreted as a buildup and the decrease as decay of the excitonic population. Recent microscopic calculations, however, have shown that even without an incoherent excitonic population, pure plasma decay yields photoluminescence peaked at the is exciton resonance.; Experimental time-resolved photoluminescence spectra are taken across a large region of the parameter space of carrier density and lattice temperature. They are compared to the expected thermal equilibrium spectra, calculated from nonlinear absorption measurements taken under identical conditions. Under none of the experimentally explored parameters is the is emission as bright as expected for thermal equilibrium.; To distinguish excitonic and plasma contributions, the deviations from thermal equilibrium at the is exciton resonance are then analyzed using a microscopic calculation. The dipole moment is adjusted to reproduce the excitonic binding energy and oscillator strength of the samples under investigation. The carrier densities and carrier temperatures are determined experimentally; no free fit parameters are necessary.; The differences between experimental values and pure plasma calculation are explained with the presence of an incoherent excitonic population. Although at first the emission spectra under all conditions do not vary significantly, a more detailed analysis reveals that the sources of the photoluminescence can be either predominantly excitonic or plasma. For low temperatures and low densities the excitonic emission is extremely sensitive to even minute exciton populations making it possible to extract a phase diagram for incoherent excitonic populations. The maximum contribution of bright excitons is found at intermediate densities and low lattice temperatures; the absolute number of bright excitons is tiny, less than 0.04% of the total carrier density. However, it is not possible to determine the total number of bright and dark exciton by using photoluminescence. |
