High-accuracy determination of the dependence of the photoluminescence emission energy on alloy composition in AlxGa1−xAs films

Abstract

In an effort to improve the accuracy of photoluminescence (PL) measurements of the Al mole fraction (x) of AlxGa1−xAs alloys, the PL peak emission energy, EPL,peak, was measured at room temperature for molecular-beam epitaxy-grown AlxGa1−xAs films with 0⩽x<0.37, and correlated with independent measurements of x by in situ reflective high-energy electron diffraction (RHEED) and also by ex situ wavelength-dispersive x-ray spectroscopy in an electron microprobe analyzer (WDS/EMPA). The measurement uncertainty of EPL,peak was minimized through the following procedures: Accurate calibration of the photon energy (or wavelength) scale, correction of the measured spectra for the spectrometer response function, fitting the data with a well-chosen line shape function, and compensation for the effect of ambient temperature drift. With these procedures, the 2σ measurement uncertainty of EPL,peak was of the order 5×10−4 eV for most samples. From correlation of the PL and WDS/EMPA composition data, the slope ∂EPL,peak/∂x near room temperature was determined to be ∂EPL,peak/∂x=(1.4017±0.0090 eV)−[(2.71±0.97)×10−4 eV/K](T−298.3 K). Correlation with the RHEED data gave the same result within measurement uncertainty. Previously published measurements of ∂EPL,peak/∂x were reviewed and compared with the present study. The results of T. F. Kuech et al. [Appl. Phys. Lett. 51, 505 (1987)], based on nuclear resonant reaction analysis of the Al mole fraction, were found to be in good agreement with the present study after the addition of a correction term to account for the sample temperature difference (T=2 K for Kuech et al., T=298 K for the present study).