Invalidation of the acquisition of internal quantum efficiency using temperature-dependent photoluminescence in InGaN quantum wells with high threading dislocation density

Abstract

Temperature-dependent photoluminescence (TDPL) is mostly employed to evaluate the internal quantum efficiency (IQE) of light-emitting semiconductors. The key assumption of this method is that the IQE is 100% under low temperature (LT) (e.g. 10 K), which is often considered to be reasonable without any verification. This may lead to an obvious contradiction between a considerable IQE value and poor emission intensity, especially when there exists a high threading dislocation density. In this paper, the power-dependent photoluminescence (PDPL), TDPL, and time-resolved photoluminescence are carried out on three InGaN single-quantum-well samples with diverse threading dislocation densities to obtain the IQE and the recombination coefficients. It is found that the IQE at LT must be confirmed by PDPL and used to calibrate the TDPL values. Then, the IQEs obtained from different measurements are in good agreement under the same excitation conditions. A phenomenological model is proposed to explain the evolution of IQE and recombination coefficients with threading dislocation density. This work indicates that the commonly used TDPL is not reliable, as the hypothesis of negligible nonradiative recombination at LT is proven to not be valid for InGaN quantum well samples, especially those with high threading dislocation density (>1010 cm−2).