Recombination Rate Analysis of InGaN-Based Red-Emitting Light-Emitting Diodes

Abstract

The recombination rates are measured and analyzed for red-emitting InGaN light-emitting diodes (LEDs) to better understand the factors that limit their efficiency. InGaN/AlGaN/GaN multiple quantum well (MQWs) are grown with x≥0.28 in the In <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">x</sub> Ga <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-x</sub> N quantum well. The Al <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">y</sub> Ga <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-y</sub> N interlayers (ILs) with high Al-content (y>0.8) are employed because they result in smoother surfaces with smaller V-pits and higher photoluminescence efficiency. The IL-MQWs are formed on GaN and In <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">z</sub> Ga <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-z</sub> N/GaN superlattice (SL) underlayers (ULs) with <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">z</i> = 0.015, 0.025, and 0.065. Differences in <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">B</i> coefficients (radiative recombination) within this set result from changes in wavefunction overlap caused by differences in layer thickness and composition in the IL-MQW. IL-MQWs grown on SL-ULs have <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">A</i> coefficients (Shockley-Reed-Hall recombination) that are lower than expected, indicating that the SL-ULs help reduce defect formation. Compared to shorter wavelength InGaN-based LEDs, the <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">B</i> coefficients are ~100 times lower due to lower wavefunction overlap. A and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</i> coefficients are higher because of a higher number of defects.