News Article
Don’t blame Auger for LED droop
Scientists say they have shown that neither direct nor indirect Auger recombination are the primary cause of droop in indium gallium nitride quantum wells.
Researchers at Technische Universität Braunschweig, Germany, say they have pinned down the Augerre combination coefficient CinInGaN quantum well structures. With C=(1.8±0.2)x10-31cm6/s, the result comes out much lower than previous experimental estimates.
The researchers say that this quantitatively measured Auger coefficient is very likely due to indirect (phonon-oralloy-assisted) Auger processes rather than direct Auger processes.
But from the small magnitude of the Auger coefficient (in relation to the radiative coefficient, which is known from independent measurements), the team, led by Andreas Hangleiter, say they can definitely conclude that neither direct nor indirect Auger recombination is the primary cause of droop in InGaN LEDs. Even though Auger recombination does qualitatively produce a droop-like behaviour, it does so only at much higher current densities than observed in LEDs.
Typical experimental internal quantum efficiency of a green LED compared to calculation based on the new Auger results showing that such a small Auger coefficient can not explain the droop.
Understanding the cause of droop is important because it limits the current density applicable to LEDs and thus the amount of light that can be generated with a single LED chip.
Hangleiter argues that if Auger recombination was the primary cause of droop, the only way out would be a reduction of the carrier density in the LED, as the Auger recombination rate scales with the carrier density cubed. Given the small Auger coefficient found here, Auger recombination is only a minor contribution to the droop. This means that other mechanisms like carrier overflow or defect-related non-radiative recombination must be reconsidered.
What’s more, Auger recombination also has a significant impact on laser diodes. It may contribute to the threshold current density and may lead to different design rules for laser diodes. Even the relatively small Auger coefficient found in this work means that non-radiative Auger processes account for 20-50% of the threshold current, even in the best nitride laser diodes.
The results obtained by the team at Braunschweig were based on optical gain spectra of InGaN-based laser structures. Optical gain spectra provide the unique opportunity to determine carrier density; their spectral width is directly given by the Fermi energies, which can be converted into carrier density. From the dependence of the non-radiative rate on carrier density, the Auger coefficient and the defect recombination coefficient can then be determined.
The researchers claim that this is in stark contrast to other attempts to investigate Auger recombination, where neither the non-radiative rate nor the carrier density are known, but are the result of a fitting procedure based on a very much simplified model.
Given these new results, Hangleiter says the droop in LEDs can no longer be explained by Auger recombination.
Further details of this work have been published in the paper “Auger recombination in GaInN/GaN quantum well laser structures”, by M. Brendel et al, Appl. Phys. Lett. 99, 031106 (2011). DOI:10.1063/1.3614557
