Osram sides with Lumileds' droop theory
Osram Opto Semiconductors Matthias Sabathil captured the key debate at the International Workshop on Nitride Semiconductors (IWNS) in his talk's opening remark: "Auger or not Auger? That is the question."
This comment referred to the competing theories for explaining the highly controversial phenomenon of LED droop, the decline in external quantum efficiency (EQE) at higher drive currents.
US LED maker Philips Lumileds claims that Auger recombination is the dominant cause of droop. But others, notably Fred Schubert's team at Renssalaear Polytechnic Institute, believe that the blame lies with carrier transport effects such as electron leakage.
Osram took its place alongside Lumileds in the droop discussion at IWNS in Montreux, Switzerland, on 9 October. The leading candidate for droop is phonon-assisted Auger recombination, says Sabathil.
German lighting giant Osram, which is collaborating with theorists from the University of Arizona and Philipps University Marburg, came to this conclusion after ruling out other potential causes of droop.
Sabathil said that defects can't explain droop, because LEDs with threading dislocation densities of 10 7cm-2 and 10 9cm-2 have very similar external quantum efficiencies over a wide range of drive currents.
Polarization charges also fail to account for droop, according to Sabathil. This can be ruled out by considering the non-polar LEDs fabricated at the University of Santa Barbara, California, which also suffer from declining EQEs at higher drive currents.
Sabathil explained that Osram has eliminated carrier injection and leakage through comparisons of electroluminescence and photoluminescence emission from a 530 nm single quantum well. Optical excitation at 405 nm ensured that carriers were only generated within the wells.
Electroluminescence and photoluminescence spectra were very similar, and Sabathil explained that this proves that the cause of droop is internal. And it doesn't depend on temperature either, according to this collaboration's measurements, so thermally activated leakage can also be ruled out.
The team went on to determine that the loss at high drive currents is proportional to the cube of the carrier density. This type of scaling law hints that the cause is an Auger-like process "“ which involves the interaction of an electron and hole with a third carrier.
From experiment to theoryThe precise nature of this process was discussed in the following talk by Bernard Pasenow from Philipps University Marburg.
Pasenow explained that the theorists within the team, which includes Jorg Hader from the University of Arizona, have ruled out the direct Auger process as the cause of droop (see Simulations question Lumileds droop theory for details).
Phonon-assisted Auger is now the leading candidate, and Pasenow revealed that this process can produce losses that are significantly larger than those resulting from direct Auger recombination.
However, he admitted that the work was still at an early stage, and that the theorists need to study the effects of material properties and phonon coupling.
