Avoiding arbitrarily wrong microluminescence statistics due to a non-quantitatively calibrated setup

Abstract

Photoluminescence (PL) measurements on the normalμm scale are a powerful tool to examine thin film semiconductors with respect to lateral inhomogeneities of their opto‐electronic properties. For practical reasons, the determination of the standard deviation of the laterally resolved splitting of quasi‐Fermi levels (QFL) is often performed with an experimental setup that only has been calibrated spectrally but not with respect to the quantitatively emitted photon flux. We show that the omission of the absolute PL photon flux can lead to arbitrarily wrong distributions of the QFL‐splitting and correlations between the QFL‐splitting and other opto‐electronic properties. For a better understanding, we present a algebraical formalism of the dependence on the absolute calibration and demonstrate the effect of a wrong calibration using two different thin film absorbers. Finally, a method to estimate the true statistics with sufficient accuracy is presented.